An emotion classification machine learning model that fuses pupillometry and facial recognition.
Explore the docs »
View Demo
·
Report Bug
·
Request Feature
Table of Contents
This project continues the research of EmoPupil, a machine learning model that uses pupillometry data to recognize emotions. This project aims to build upon this by adding in a facial expression modality that will give the model more information to work with, and therefore the potential for better emotion classification accuracy.
This figure demonstrates the training pass of our machine learning model.
This figure demonstrates the testing pass of our machine learning model.
- Data collection
- Data processing
- Machine learning model
- Model training and hyperparameter tuning
See the open issues for a full list of proposed features (and known issues).